For many years, academics have had instincts about what topics in psychology are subject to the strongest taboos, but only recently has an attempt been made to determine the answer empirically. In a study published in 2024, a team of researchers surveyed 470 psychology faculty at 133 American universities, asking them their views on 10 taboo conclusions that had been nominated by their peers. Although this was not the only metric examined, the hypothesis that those surveyed were most inclined to discourage other academics from studying was the following: “Genetic differences explain non-trivial (10% or more) variance in race differences in intelligence test scores” (Clark et al., 2025). This hypothesis also had one of the strongest correlations between self-censorship and belief in the hypothesis’s veracity, with those who believed it to be correct being the most afraid to express their views about it.

This paper will not take a position on whether the hypothesis is correct or incorrect, but will argue that such a taboo is socially harmful, and that it has worsened the problem of racial discrimination in American society.

Thus wrote Thomas Jefferson in the Declaration of Independence. This statement is among the founding principles of the United States, but there are competing interpretations as to what it means.

One interpretation is that all human beings are equally entitled to rights such as freedom of speech, freedom of religion, the right to own property, and the right to receive a fair trial when charged with a crime. It follows that the duty of governments is to protect these rights of all individuals, without showing favoritism based on race, class, gender, or religion. From this perspective, the statement “all men are created equal” is silent about whether humans are equal in their abilities or tendencies, because a person’s rights do not depend on their personality or skills.

Historically, this is the meaning that Jefferson intended, as his other writings indicate that he did not believe individuals or groups to have identical abilities. In a letter to John Adams, Jefferson wrote that there exists “a natural aristocracy among men”, and that “The grounds of this are virtue and talents” (Coffman, 2012, p. 261). In his 1785 book Notes on the State of Virginia, Jefferson also expressed the opinion that the distribution of intellectual abilities varies among racial groups (Jefferson, [1785] 1832, pp. 144–51). A century before the first IQ tests, this was a highly speculative conclusion, and Jefferson’s views on race and intelligence were expressed using crude terms such as “superior” and “inferior” that were typical of his time. However, in a letter to Henri Grégoire, Jefferson emphasized that his views on race differences by no means diminished his belief that all are deserving of equal rights: “whatever be their degree of talent it is no measure of their rights. Because Sir Isaac Newton was superior to others in understanding, he was not therefore lord of the person or property of others” (Ford, 1898, pp. 246–47).

The other possible interpretation of the statement “all men are created equal” is that all human beings—or demographic groups of human beings—possess equal abilities and tendencies in socially valued traits, and therefore all are deserving of approximately equal life outcomes. In modern times, the term equality usually refers to members of all groups being entitled to equal treatment, while equity refers to all groups of people having roughly equal outcomes. The concept of equity is famously explained by Ibram X. Kendi in his influential book How to Be an Antiracist:

A central assumption of Kendi’s perspective is that racial inequities have only one possible cause. He explains, “Antiracist ideas argue that racist policies are the cause of racial inequities” (Kendi, 2019a, p. 20). In an article published as part of a Politico symposium, Kendi (2019b) argued that the United States should amend its Constitution with a statement of these two principles of the drive for equity: “Racial inequity is evidence of racist policy and the different racial groups are equals.” In this context, the term “equal” is not being used in Jeffersonian sense—deserving of equal rights—because Kendi has made it clear that achieving equity will require discriminating based on race: “The only remedy to racist discrimination is antiracist discrimination. The only remedy to past discrimination is present discrimination. The only remedy to present discrimination is future discrimination” (Kendi, 2019a, p. 19).

If all demographic groups of humans deserve approximately equal average outcomes, and therefore “antiracist discrimination” is necessary to achieve racial equity, this cannot coexist with Jefferson’s original premise that everyone deserves equal treatment under the law. Kendi’s book acknowledges that these two goals are incompatible, describing “the construct of race neutrality” as the greatest enemy to achieving racial equity (Kendi, 2019a, p. 20). Although this might seem an extreme conclusion, it is difficult to avoid if one accepts the premise that all racial disparities are directly or indirectly caused by racism. If Kendi’s premise is correct, it would mean that racism is shaping the demographics of every area of American society where racial disparities exist, and such a situation could reasonably be regarded as a moral emergency. Accepting this premise also would mean that, theoretically, it should be possible to reform society in a way that makes all racial disparities disappear.

This is what I call the racial equality paradox: if one is committed to giving equal treatment to all individuals regardless of race, it is necessary to reject Kendi’s view that all racial disparities are caused by racism (and do not reflect any difference in the underlying distribution of preferences or abilities), as well as the resulting goal of equality of outcomes. This does not mean abandoning the goal of eliminating racial prejudice where it exists, nor does it mean giving up on reducing the size of disparities when that can be accomplished without engaging in racial discrimination. However, it means accepting that eliminating racial disparities altogether is not a realistic goal, at least not without some as-yet-undiscovered intervention unlike anything that has been attempted up to this point. It also means understanding that in cases where these disparities continue to exist even after visible forms of discrimination have been eradicated, their continued existence does not justify blaming or discriminating against any group of people.

Although the term “equality paradox” with respect to race is coined in this paper, a related concept called the “gender equality paradox” is already well known. The latter term refers to the widely observed phenomenon that gender differences in career choices tend to be more pronounced in countries with a greater degree of social and political equality between the genders, with fewer women tending to choose STEM jobs in the societies where women have the most freedom (West et al., 2019). The focus here will be on the equality paradox with respect to race, but the broader principle also should be noted: with respect to both race and gender, there is a tradeoff between equality of outcomes and equality of rights.

Kendi’s worldview was one of the guiding principles of the Biden administration. Vice President Kamala Harris (2022) explained her understanding of this principle at an event for the Oakland Generation Fund, a program of financial assistance for children of low-income families. “So equity, as a concept, says: Recognize that everyone has the same capacity, but in order for them to have equal opportunity to reach that capacity, we must pay attention to this issue of equity if we are to expect and allow people to compete on equal footing.” In his remarks about signing an infrastructure bill into law, President Joe Biden (2022) put it more succinctly: “Talent is equally distributed, but opportunity is not.” On the basis of this worldview, Biden (2021) issued an executive order to advance “an ambitious whole-of-government equity agenda.”

The ways that this executive order shaped the Biden administration’s policies were most evident in their COVID relief initiative known as the “American Rescue Plan”, which prioritized funding by both race and gender. This program provided financial relief to farmers only if they belonged to “socially disadvantaged” groups. This category was defined as farmers who were African American, Native American or Alaskan native, Hispanic, Asian American or Pacific Islander, while White farmers were not eligible for the benefits (Gerstmann, 2021b). A federal judge eventually halted this program on the grounds that it constituted illegal racial discrimination, explaining that “Plaintiffs are excluded from the program based on their race and are thus experiencing discrimination at the hands of their government” (Griesbach, 2021).

Another program established by the American Rescue Plan, known as the Restaurant Revitalization Fund, prioritized benefits to minority-owned restaurants (Gerstmann, 2021a). According to this fund’s criteria, only certain specific national origins conferred “minority” status—for example, individuals with ancestry from Pakistan and India qualified for the special treatment, while those with ancestry from Afghanistan, Iran, and Iraq did not. Like the racially exclusive benefits to farmers, the racial component of this program was halted by a federal judge, who commented that “The stark realities of the Small Business Administration’s racial gerrymandering are inescapable” (Norris, 2021). Following these court rulings, the Biden administration implemented programs of racial preferences in two further pieces of legislation: the Infrastructure Investment and Jobs Act and the CHIPS act, both of which prioritized grants for minority-owned businesses (Chandler et al., 2024). Among the provisions of these bills was to expand the scope and funding of the Minority Business Development Agency, which went on to be the subject of a lawsuit alleging that it was engaging in illegal racial discrimination (Mark, 2023). As in the other cases, the court eventually found this agency’s system of racial preferences to be unconstitutional (United States Congress, 2024).

The policies for which this agency was sued were part of a larger constellation of racial preferences, mostly at the state level, that exist in benefits for businesses. Most U.S. states have adopted programs whereby businesses may obtain various special benefits by becoming certified as MBE (minority business enterprise) or MWBE (minority and women-owned business enterprise) (Butkovich, 2021), including some programs that predate the Biden presidency. The precise requirements for a company to obtain this certification vary by state, but a typical requirement from the state of California is for a company to be “at least 51% owned by one or more minority individuals, and whose management and daily operations are controlled by one or more of those individuals” (California Public Utilities Commission, 2025). Among the benefits that California provides for MBEs is a statutory goal to award these businesses “15% of all contracts awarded by any state agency, department, officer, or other state governmental entity” (Arambula, 2007). Unlike the federal racial policies implemented under Biden, these state-level policies are still commonplace. California’s MBE program is active as of 2025, and in June 2025, the state government of New York extended its own MWBE program through 2028 (Hancock Estabrook, 2025).

Below the federal and state level, programs of racial discrimination in the name of equity also have been adopted in local governments. It is impossible to list every such policy, but keeping the spotlight on California, a typical example is a pair of racially exclusive financial stimulus programs adopted in 2021 by Marin County and Oakland. To be eligible for the cash benefits under the Oakland program, a person was required to have at least one child under 18, be earning no more than 50% of the area’s median income, and be BIPOC (Black, Indigenous, and/or people of color) (Batchelor, 2021). The Marin County program provided cash grants to low-income women who were raising at least one child, but only if they were “women of color” (Cabanatuan, 2021). The Marin County program was run by a non-profit organization called the Marin Community Foundation, with financial support from the Marin County Board of Supervisors, while the Oakland program was run by the county government itself (Britschgi, 2021). The Marin County and Oakland programs ended in 2024 and 2025 respectively, not due to any legal challenge, but because they were intended as experimental and the experiment had run its course (Halstead, 2024; Ravani, 2025).4

This discussion has been mainly limited to discrimination in the name of equity by the U.S. public sector, as similar cases in the private sector are too numerous to provide a complete overview.5 A brief selection of public sector examples from other U.S. states includes a state-run homebuying program in Washington state that excluded applicants based on their race; state guidelines for treating COVID-19 in New York that prioritized treatments by race (Friedersdorf, 2024); statewide racial preferences in Vermont’s COVID-19 vaccination program (Galewitz, 2021); and a pandemic relief grant program run by the state of Oregon that was offered only to residents who identify as Black (Eligon, 2021). A separate but related outcome, previously covered in JCI by Abbot et al. (2023), is the rejection of merit-based standards by scientific journals and institutions, on the assumption that racial disparities in STEM fields demonstrate that merit-based standards perpetuate systemic racism. Fields discussed by Abbot et al. that have been impacted by this change include medicine, psychology, physics, chemistry, and global health.

The Universal Declaration of Human Rights, an international document adopted by the United Nations (United Nations General Assembly, 1949), states that “All are equal before the law”, and that “Everyone has the right of equal access to public service in his country.” This document is among the cornerstones of modern liberalism, and is reflected in human rights laws and national constitutions around the world (Tomuschat, 2017). In the United States, racial discrimination in employment, public accommodations, and federally funded programs also has been prohibited for over sixty years by the Civil Rights Act of 1964 (National Archives, 2022). Similar principles are enshrined in several state constitutions, including a provision in the state constitution of California that forbids both the state and any city or county within it from providing preferential treatment based on “race, sex, color, ethnicity, or national origin in the operation of public employment, public education, or public contracting” (California State Legislature, 2019). Marin County and Oakland’s programs of race-based benefits were adopted less than six months after California’s voters had rejected a proposal to repeal this provision of their state constitution (Friedersdorf, 2020).6 By abandoning what had previously seemed a well-established moral and legal principle, the Biden administration, as well as American state and local governments, have exhibited a fundamental shift with respect to the role of governments in a society.

The rejection of this tenet of liberalism, and the belief that governments must provide preferential treatment based on race, was an inevitable endpoint of discourse about racial disparities in the United States. As programs such as the No Child Left Behind Act have not succeeded at eliminating the disparities, proposals for overt racial preferences have grown in popularity because they are presumed to be the only remaining solution (Howard & Navarro, 2016). The assumption is that as large racial disparities have persisted despite heroic efforts to eliminate them, simple explanations based on factors such as the quality of public schools have gradually become less plausible, and the disparities must instead be a product of subtle forms of racism. Although racial bias in the United States undoubtedly still exists, these arguments take it for granted that racist social systems are the cause of ongoing racial disparities, without attempting to empirically establish a causal connection.

Aside from Kendi’s How to be an Antiracist, probably the most influential contemporary antiracist book is Robin DiAngelo’s White Fragility. Like Kendi, DiAngelo (2018, pp. 22–23) makes this assumption: “Many whites see racism as a thing of the past, and of course, we are well served not to acknowledge it in the present. Yet racial disparity between whites and people of color continues to exist in every institution across society, and in many cases is increasing rather than decreasing.” A third major antiracist text, which presents itself as an overview of the critical race theory (CRT) perspective, is the fourth edition of the textbook Critical Race Theory: An Introduction by Richard Delgado and Jean Stefancic. Delgado and Stefancic (2023, pp. 11–13) argue that despite the disappearance of overt manifestations of racism such as lynching, the racist nature of American society is demonstrated by the continued existence of racial disparities in education, socio-economic status, occupational status, incarceration, and health outcomes. In a section titled “Critique of liberalism”, Delgado and Stefancic’s book goes on to argue:

These developments are the ironic background against which research about group differences in average ability is accused of supporting the racialized structure of society. As the area of study has been accused of causing this social harm, U.S. federal, state, and local governments were simultaneously adopting the “aggressive, color-conscious efforts” advocated by antiracist authors and publications, justified on the belief that racial disparities cannot have any other explanation besides racism. That research about group differences becoming taboo would increase the importance American society places on racial categories, instead of diminishing it, also was not entirely unpredictable. Over a decade before these extreme measures became commonplace, in a book chapter about social consequences of group differences, University of Delaware psychologist Linda Gottfredson warned that results such as these were the cost of assuming that all racial disparities in life outcomes must instead reflect the presence or absence of social privilege. (Gottfredson’s summary uses the term “inequality” for what is nowadays called “inequity”, because it was written before the term “inequity” became the accepted term for racial disparities.) Gottfredson wrote:

At the time when the book containing this quote was published, Gottfredson’s summary was an extremely pessimistic description of American society, because in 2006 concepts such as White privilege, White fragility, and the field of Whiteness studies had not yet entered the intellectual mainstream. Her summary is best understood not as a literal description of race relations in the early 2000s, but as a warning about the eventual results that would occur if society refused to discuss group differences and their social consequences. In general, the ability to make accurate predictions is regarded as a key characteristic of a successful theory or research program (Lakatos, 1974; Kane, 2016). While most predictions made by research about group differences have been centered on genetic or psychometric data, Gottfredson’s was an example of such a prediction with a broader scope, albeit one whose eventual fulfillment was very unfortunate.

The majority of public discussion about racial equity has focused on groups that are under-represented in socially valued roles relative to their share of the population. But the knife cuts both ways: if racial inequities are a sign of injustice and require racial preferences to address them, then that is also true when minority groups are over-represented in socially valued roles. Despite its high-minded justification, in practice this type of discrimination has often resembled White resentment directed at more successful groups. It has especially affected people of East Asian and Ashkenazi Jewish descent, two groups whose average IQ scores are higher than the average score of people of non-Ashkenazi European descent (Winegard et al., 2020).

When all demographic groups are assumed to have equal average cognitive ability, a separate form of racial discrimination that tends to result is the practice of race-norming IQ scores in the American criminal justice system. The background to this practice is the 2002 Supreme Court case Atkins v. Virginia, which found that executing a person with an intellectual disability violates the constitutional prohibition against cruel and unusual punishment (United States Supreme Court, 2002). Intellectual disability is diagnosed primarily using intelligence tests, based on an individual scoring at least two standard deviations below average (that is, having an IQ score of 70 or lower), with an error margin of 5 points (American Psychiatric Association, 2013, p. 37).

The concept of race-norming is based on the assumption that group differences in average IQ scores do not reflect a real difference in average ability, and that therefore when determining whether a Black or Hispanic defendant may receive the death penalty or whether that is prohibited by the Atkins ruling, their IQ must be adjusted upwards to compensate for the racial bias that is allegedly depressing the score (Sanger, 2015). In other words, when this practice is in effect, if a Black and White defendant both are convicted of identical murders and both have an IQ that is measured at 63, the Black defendant will be eligible for the death penalty but the White defendant will not. The White defendant’s IQ score would classify him as intellectually disabled, and thus exempt from the death penalty according to Atkins. On the other hand, on the assumption that IQ tests underestimate the ability of African Americans, the Black defendant would have his IQ adjusted upwards by 10 to 15 points, making him eligible for it.

Instances of race-norming have usually relied on the “clinical judgment” of expert witnesses called by the prosecution, most often without providing any empirical evidence to support these ethnic adjustments of test scores. A well-known example of this practice was the case of Ramiro Hernandez, a Mexican American defendant whose IQ had been variously measured at 52, 54, and 57 (62 according to some sources). If Mr. Hernandez had been White, these scores would have made him ineligible for the death penalty according to the Atkins ruling. However, a prosecution witness successfully argued (without ever meeting Hernandez personally) that due to the defendant’s Mexican ancestry, his IQ score must be adjusted upwards to 70. As that score is on the edge of what is considered intellectually disabled, Mr. Hernandez was sentenced to death (Shapiro et al., 2019).

Hernandez’s case, and the broader practice of race-norming, received attention when a lawsuit to prevent his execution was appealed to the U.S. Supreme Court in 2014. The brief presented to the Supreme Court described many other cases in which prosecutors had argued—sometimes successfully—that Black and Hispanic defendants should receive the death penalty despite having IQ scores in the range where it normally is prohibited. These arguments had relied on such claims as that IQ tests are “culturally and racially biased”, or that they “tend to underestimate particularly the intelligence of African-Americans” (Young & Blondel, 2014). However, the Supreme Court declined to hear Hernandez’s case, and he subsequently was executed (Sanger, 2015).

The reasoning behind the Atkins decision is that individuals with intellectual disabilities have reduced ability to learn, to understand the world, and to control their impulses. While such a disability does not exempt a person from criminal penalties altogether, it does diminish their culpability (Grunewald, 2002). IQ tests function the same way and measure the same abilities regardless of the race of the person tested (Warne, 2020, p. 255), and it is well established in differential psychology research that IQ tests are not culturally biased against any race or ethnicity (Warne, 2020, pp. 91–92; Reynolds et al., 2021, pp. 573–613; Kalat, 2022, pp. 299–300). To put it another way, a defendant having an IQ below 60 represents the same degree of diminished culpability regardless of his or her race. The assumption that this score has a different meaning if a defendant is Black or Hispanic, and that this indicates a greater criminal culpability than if they were White, is based on a misunderstanding of group differences.

A paper by Robert M. Sanger (2015), which provides the most complete overview of the practice of race-norming, explains its appeal: “A superficial attraction of the prosecution’s argument is its appeal to a sort of benign political correctness. Chiefly, the argument is that to not adjust African Americans’ or Latina/os’ scores would be to make the racist assumption that members of each ethnic group are genetically inferior.” The adverse effect of this attitude is a further example of a pattern that also includes notions such as “Jewish privilege,” and that Asian Americans are deficient in personality. The shared pattern is that when concerns for racial justice result in the assumption that all demographic groups have equal average ability, this attitude tends to backfire in unpredictable ways, and to instead support new, more destructive forms of prejudice or discrimination against minority groups.

The reason for this outcome can be explained by a principle that James Flynn, the eponym of the Flynn Effect, warned about in a letter to Nature. Paraphrasing John Stuart Mill, Flynn (2009) warned that when an area of research is considered off-limits, it is impossible to know every application that could have emerged from the eschewed knowledge, and only someone who considers themselves infallible could claim to know that cutting off such a line of inquiry is for the best. In this area, ignorance about group differences has unpredictably supported a trend whereby in the United States, minority defendants are more likely than White defendants to be executed for similar crimes—a trend that advocates of racial justice commonly decry (Keys & Maratea, 2016; Maratea, 2019; American Civil Liberties Union, 2003). For those who are legitimately concerned about racial justice, rather than only signaling their virtue with performative gestures, the ignorance that is contributing to this problem should be a major concern.

When psychologists and behavioral geneticists study the various factors that can affect an IQ score, and the score on other similar tests such as the Armed Forces Qualification Test and the Woodcock-Johnson test, they are studying influences on far more than the test score itself. The reason cognitive ability tests are widely used in these fields is that they are the most effective known way of measuring a set of cognitive skills that affect many areas of life (Deary, 2000, pp. 21–25). Plomin and Asbury (2014, p. 89) summarize the data in this way: “It is not overstating the case to say that IQ is the best single behavioral predictor of future success that we currently have.”

A chapter by Tarmo Strenze (2015) in the Handbook of Intelligence: Evolutionary Theory, Historical Perspective, and Current Concepts reviews several meta-analyses which have found that success in various areas correlates with IQ scores. Among these are a correlation of 0.58 with academic performance in primary education, a correlation with 0.53 with measures of job performance (as rated by supervisors), a correlation of 0.43 with occupational attainment, and a correlation of 0.20 with income. These data are shown in the graph below.

To ensure that it measured the influence of cognitive ability on income, rather than the influence of socio-economic status on IQ, the meta-analysis examining the relation between these variables (Strenze, 2007) used only longitudinal studies, finding that IQ measured early in life predicts income in adulthood. This meta-analysis also found that IQ is a stronger predictor of income than parental socio-economic status or parental education. The fourth edition of the textbook Individual Differences and Personality by Michael C. Ashton (2022, pp. 267–76) discusses two other such correlations: individuals with higher IQ scores tend to be healthier and have longer lives, and also are less likely to commit crimes, even when controlling for socioeconomic status. Consequently, when researchers examine influences on human cognitive ability (as measured by intelligence tests), they are indirectly examining influences on these various other life outcomes as well.

As stated in the previous section, it is widely accepted in differential psychology that intelligence tests do not show cultural bias. This conclusion has been supported in reports published both by the National Academy of Sciences and by the American Psychological Association. The better-known of these two reports, Intelligence: Knowns and Unknowns, was commissioned by the American Psychological Association in response to the controversy surrounding Herrnstein and Murray’s book The Bell Curve in the 1990s. Although the APA report did not support Herrnstein and Murray’s argument that group differences in average test scores include a genetic component, it agreed that “Considered as predictors of future performance, the tests do not seem to be biased against African Americans” (Neisser et al., 1996).10 The earlier report, published in the 1980s by a committee appointed by the National Academy of Sciences, reached a similar conclusion:

When differences in average cognitive test scores exist between racial groups, and when the tests also predict life outcomes equally well for all groups, it is inevitable that group differences in average test scores will be reflected in areas of life where the abilities measured by these tests make a difference. A few studies have confirmed this to be the case. A pair of studies by Kevin Beaver et al. (2013a, 2013b) found that the relationship between lower IQ and greater risk of criminal involvement exists across multiple race and gender groups, and that the disparity in arrest and incarceration rates between Black and White American men disappears when controlling for both IQ and self-reported lifetime violence. (However, this does not rule out the existence of other forms of racial discrimination in the American criminal justice system, such as the one described in the preceding section.) Several meta-analyses also have examined group differences in average measures of job performance, finding that these differences are larger in more cognitively demanding tasks (Roth et al., 2001) and that they are about the same size in both subjective and objective measures of job performance (Roth et al., 2003), and replicating both of these results with a third, larger meta-analysis (McKay & McDaniel, 2006).

In a discussion of these various lines of data, Earl B. Hunt’s 2011 textbook Human Intelligence concludes that racial disparities in job performance are “about what would be predicted from the gaps in cognitive test scores and the correlation between test scores and performance.” With respect to studies that examined the relation between test scores and income, Hunt also concludes that the relation between those two variables is “essentially identical” for all of the racial groups examined, meaning that a person with a given test score will, on average, earn approximately the same income regardless of their race (Hunt, 2011, pp. 420–21). This is not a minor source—psychologists regard Hunt’s textbook as one of the most authoritative books published about human intelligence in the early 21st century (Deary, 2012; Johnson, 2016). When a racial disparity is of the size that would be predicted from group differences in average IQ scores, this is sometimes known as “IQ commensurability” (Gordon, 1997).

When it is accurately understood, this body of research undermines both the justification for modern forms of racial discrimination in the name of equity, and older forms of discrimination against disadvantaged groups. Supporters of race research such as Russell Warne (2020, pp. 326–27) and Charles Murray (2021, p. 39) have emphasized that an individual’s abilities cannot be determined from his or her group average, and that therefore all people must be judged based on their individual merits, not based on their racial identities. Warne (2020, p. 316) also discusses the importance of making society as accessible as possible to individuals with low cognitive ability. A similar principle is widely accepted with respect to disadvantages such as blindness or autism, but public ignorance about intelligence research has limited the efforts to assist people whose cognitive ability is below average, when it is not low enough to qualify as an intellectual disability.

These moral principles have not been enough for opponents of race research such as Eric Turkheimer, who argues (2024, pp. 143–44) that contrary to claims about the research not supporting racial prejudice or discrimination, any argument that one group is predisposed to be less intelligent than another is inherently derogatory. However, Turkheimer also accepts that equality in the moral sense is a more fundamental ideal than any hypothesis about the distribution of ability, writing (2024, p. 7) that “For all his imperfection and hypocrisy, Jefferson’s famous sentence meant that people are morally and politically equal despite their obvious differences.” The most important ethical question about race research (as about many other areas of study) is not how it relates to this ideal in isolation, but whether the conclusions of race research are an improvement over what inevitably tends to be concluded in its absence. On the assumption that racial disparities are mostly or entirely due to racist social structures, antiracist scholarship has concluded that all White people who are not actively antiracist are guilty of participating in an unjust social system which benefits them at other groups’ expense. This purported form of racial culpability is known as “white complicity” (Applebaum, 2010; Daum, 2020), and it amounts to a race-based moral judgment of (nearly) all members of a group. Like Kendi’s (2019a) rejection of “the construct of race neutrality”, it also is a conclusion that is difficult to avoid if one accepts its premise about the cause of racial disparities.

There is an ethical imperative to challenge this worldview, because it is the basis for widespread racial discrimination in modern times, undermining one of the central principles of liberal democracies (Forgas, 2025). The assumption that all groups have equal average ability is part of the intellectual foundation of recent discrimination against White Americans by the U.S. public sector, against Jewish and Asian Americans in higher education, and against Black and Hispanic Americans via the practice of race-norming. Arguments that research about group differences is socially harmful, such as those presented in Hastings Center Report and at the AABA symposium, virtually never discuss the recent social consequences of arguments similar to theirs having gained acceptance. When these arguments are made sincerely, and not as a cynical strategy to support racial preferences, they are founded primarily on superficial judgments about what research and ideas often have existed in the company of racial prejudice, rather than on any careful examination of what is or is not actually conducive to a non-discriminatory society.

This point should not be construed as denying the capacity for research about group differences to be misused by individuals on the far right, any more than discussing the value of virology research denies its potential to be misused for biological weapons. In all potentially dangerous areas of study, researchers have a duty to use their knowledge responsibly, and to do what is possible to encourage the same ethical standards in others. In a paper published in Nature, the eminent twentieth-century biologist Bernard Davis explained this situation’s key principle:

Davis (1978) is explicit that the principle he is describing applies to the potential “demonstration of heritable differences in the distribution of abilities among different ethnic groups,” but Davis himself did not name specific examples of how such research could be used for beneficial purposes. However, almost fifty years later, Bernard Davis’s conclusion with respect to this area of study is no longer merely theoretical. Instead of trying to prevent research about group differences, out of fear of its well-known capacity to be misused, it would be wiser to ensure that it is used for one particular, essential benevolent purpose. That purpose is to protect the existence of a society in which all individuals have equal rights, one in which no group is punished or disparaged because of its success, and one that recognizes a principle of liberalism described by John F. Kennedy (1963, p. 446): “All of us do not have equal talent, but all of us should have an equal opportunity to develop those talents.”